This report outlines the beginning steps taken in an integrated research effort toward the development of a methodology, and supporting systems, for transforming Ada programs, or program units, (directly) into corresponding VLSI systems. The time seems right to expect good results. The need is evident; special purpose systems should be realistic alternatives where simplicity, speed, reliability, and security ae dominant factors. Success in this research can lead to attractive options for embedded system applications. Ada programs can be regarded as ensembles of machines, one per program unit (module), which in turn may be mapped directly into corresponding VLSI structures on one or more chips with interconnecting (packet switched or other) communication nets. The research reported here is part of a five-year plan, the first year of which focuses on 'proving' the concepts through a realistic demonstration of methodology for a specific example Ada program (a silicon representation of part or all of the DoD Standard Internet Protocol, IP, initially expressed in Ada). Implicit in these objectives is the development of a set of hardware structuring paradigms (rewrite rules) whose application can ensure that transformation steps between levels of abstraction in the design process are well structured in order to preserve the integrity and, where possible, the clarity of the original Ada specification. Some paradigms, but of course not all, lead to highly efficient implementations.